Select variables for principal component analysis
From the PCA, three components were retained. These three components, as well as one additional variable, sale_pricesft_ch, defined as the percentage change in median arms-length home sale prices per sqare foot, from the period of 2014-2016 to 2017-2019.
The cluster analysis settled on a solution identifying eight distinct neighborhood types among Cleveland’s block groups.
The names given to the eight neighborhood types- from cluster “A” to “H”, correspond to the average median home sale price for the block groups in each cluster (i.e., block groups in cluster “A” have the lowest average sale price). The clusters in the first map are colored according to the same scale. In the second map, the clusters are colored by the average median household income in the block group.